Most readers of this blog will be familiar with High-Impact Practices, a set of teaching and learning practices that have been widely tested and shown to benefit students from all backgrounds. Each of the practices incorporates active learning techniques and has demonstrated impact on student retention and engagement – particularly for students from historically underrepresented groups.
Although the evidence to support these practices seems solid, we know less about how best to implement them. Each campus has its own mission, culture, and unique local context that influence efforts to scale-up educational interventions. In this Spotlight on Success, we take a closer look at how one university used an iterative process to implement a high-impact practice, and which aspects of that process contributed most to their success.
UC Riverside’s STEM Learning Communities
In the early 2000s, UC Riverside’s College of Natural and Agricultural Sciences (CNAS) decided to try and address the high DFW rates in freshman gateway math and science courses by piloting a high-impact practice: learning communities. Initially, students in the communities only had to attend a faculty discovery seminar each Fall quarter. The seminars had a negligible effect, so in 2007 an advising component was added, then co-enrollment in gateway courses, supplemental instruction (SI) from peer mentors, and lastly an early team-based research engagement.
By 2012, nearly five years into the full-blown pilot, the learning communities were starting to deliver concrete results. For example: graduation rates were up 20 percentage points, 96% of freshman were retained into their second year, and first-quarter GPAs had increased by 0.3 points among students in the learning communities. Because the pilot was set up as an experiment, CNAS had a control group with which to compare the students in the learning communities. Interested students would sign up at summer freshman orientation and then the college would select participants through a randomized lottery until the communities were full.
The learning communities were equally significant for students from historically underrepresented groups. UC Riverside has an unusually diverse first-year student population; CNAS alone is 60 percent first-generation students, 47 percent underrepresented racial/ethnic minority, 55 percent low-income and 57 percent female. Disparities in performance between underrepresented minority groups and their majority peers have narrowed considerably or disappeared. UC Riverside is also one of very few campuses that has no gap in its overall graduation rate across racial/ethnic groups.
Determinants of Success
I spoke with Dr. Michael McKibben, Divisional Dean of Student Academic Affairs at CNAS, and asked him what he thought made the program so successful. “What really generated improvement is when we put together this constellation of multi-faceted engagements,” said McKibben. “Each component on its own is marginally helpful, but when you put all the components together, you start to see a much bigger impact.” Co-enrollment in STEM gateway coursework and the SI has created a sense of community and provided students with “shared challenges” social support which, for many low-income and first-generation students, is critical for freshman year confidence and success.
Another key factor for success was the undergraduate research component. Evidence shows that STEM students who engage in undergraduate research are more likely to stay in STEM and complete their degrees. Anecdotally, engagement in the research clusters has increased students’ confidence and motivation to ask for help from faculty. “It breaks down the ‘intimidation barrier’”, says McKibben, “UC faculty are much more research-oriented than students’ high school instructors, and students might not feel comfortable approaching them initially.”
Challenges and Barriers
Getting faculty-buy in was a challenge at first, McKibben admits, because they couldn’t see the immediate impact on their first quarter freshman. Continuing to share statistics about the program’s long-term success has helped to win over faculty. The program is expensive to run, and requires huge investments in staff time and funding for research stipends. Senior leadership are on board because the program has external funding. However, CNAS is still working on securing a more permanent source of funds and integrating the program into core institutional operations. CNAS would also like to serve more students –right now, more students want to be part of the learning communities than the college has resources to support.
UC Riverside’s approach has generated a lot of attention at both the state and national level. Other colleges at UC Riverside, as well as other University of California institutions, have begun to adopt CNAS’ methods. The program and its impact on graduation rates has won national awards, from APLU and from the White House Initiative on Educational Excellence for Hispanics. CNAS hopes to leverage its success into a new pilot of second year learning communities, providing sophomores with co-curricular enrollment, career mentoring by alumni, and research skills development courses. The college also hopes to expand its Research in Science and Engineering (RISE) component, which provides rising sophomores and incoming transfer students with a $5,000 stipend to do full-time, faculty-led research for 10 weeks in the summer.
McKibben has some advice for other colleges seeking to implement learning communities in STEM:
Set up control groups and run the pilot for a number of years, so that you can see the impact of the program on graduation.
Be patient! Don’t expect that if you just put first-generation students in a room together for an hour a week that something dramatic will happen. It may take several years and tweaking a “constellation” of interventions before you start to see improvement.
There is no one-size-fits-all approach – all students respond differently to various engagement strategies. A constellation of interventions increases the odds that all students will benefit.